🔌
New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
// RUN — FITS
Can the Apple M3 Max 48GB run RNA-FrameFlow?
Yes — here's how fast and up to what context, with real measured numbers.
RNA-FrameFlow on the Apple M3 Max 48GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 636k | 8095.2-9761.9 t/s | ~202.4-244 (slow) |
| Q8 | yes | 636k | 16190.5-19523.8 t/s | ~404.8-488.1 (slow) |
| Q4 | yes | 636k | 28911.6-34863.9 t/s | ~722.8-871.6 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.
FAQ
Can the Apple M3 Max 48GB run RNA-FrameFlow?▶
Yes. At Q4 it generates ~28911.6-34863.9 tok/s and fits up to 636k context; beyond that it spills to SSD swap and slows to ~722.8-871.6 tok/s.
How much context fits?▶
At Q4, up to about 636k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.